CMAD: Cooperative Multi-Agent Diffusion via Stochastic Optimal Control

ArXi:2602.10933v2 Announce Type: replace Continuous-time generative models have achieved remarkable success in image restoration and synthesis. However, controlling the composition of multiple pre-trained models remains an open challenge. Current approaches largely treat composition as an algebraic composition of probability densities, such as via products or mixtures of experts. This perspective assumes the target distribution is known explicitly, which is almost never the case.